> Well... It is true of everything in that blog post.

That turns out not to be the case.

> I would be really interested in seeing a study that failed to find a benefit of type systems for refactoring tasks in large projects.

You're inverting the burden of proof here. People who claim that something is beneficial are the ones who have to show that it is.

Making a claim and then shouting "prove me wrong!" is not how science works. Not even computer science.